Osler-Rendu-Weber disease, or hereditary hemorrhagic telangiectasia (HHT), is an autosomal dominant condition causing recurrent epistaxis, telangiectases and visceral vascular anomalies. It is caused by a mutation in either endoglin, a transforming growth factor - receptor on human chromosome 9, or an unidentified gene on chromosome 12, or perhaps other, as yet unidentified, gene(s). Common sources of serious morbidity in HHT are pulmonary vascular arteriovenous malformation (PAVM), cerebral arteriovenous malformation (CAVM) and hepatic arteriovenous malformation (HAVM). The prevalence of PAVM in HHT is not well established, but appears to be significantly higher in families with HHT due to endoglin mutations than in those with HHT due to mutations in another gene. Several methods exist to screen for PAVMs, including contrast echocardiography, pulse oximetry, and helical computed tomography. No prospective trials have compared the sensitivity or specificity of these screening methods. While it is generally agreed that magnetic resonance imaging is the method of choice in screening for CAVMs, the prevalence of these lesions is not well established. Similarly, while Doppler ultrasound appears to be the method of choice for screening for HAVMs, the prevalence of these lesions is also not well established. No study has examined whether any modifier gene(s) exists that determine which individuals with an endoglin mutation develop PAVMs, CAVMs or HAVMs. Also, while endoglin is abundantly expressed on heart valves, it is not known whether those with endoglin mutations have increased risk for valvar anomalies. This study seeks to investigate six issues: 1) the prevalence of PAVM in a population with HHT due to endoglin mutations; 2) the relative sensitivity and sensitivity of pulse oximetry, helical CT and contrast echocardiography as screening modalities for PAVM; 3) the incidence of CAVM in a population with HHT due to endoglin mutations; 4) the incidence of HAVM in a population with HHT due to endoglin mutations; 5) the existence of modifier genes that determine which individuals with endoglin mutations develop PAVMs, CAVMs or HAVMs; and 6) the prevalence of abnormalities of the cardiac valves in those with endoglin mutations.